基于大数据条件下的管道风险评估方法思考

过去几十年,管道风险评估技术得到快速发展,但依然沿袭传统的风险评估技术发展之路,在专家风险认知、事故概率统计等基础上,利用对外在风险认知程度对评估对象进行预测分析,其核心问题是风险评估的模型是基于有限的数据基础对整个系统进行评估的理念。随着管道发展到大数据时代,在实现管道属性相关数据对齐并纳入数据库的条件下,形成了管道大数据,在很多方面具备了＂样本=总体＂的基础,使风险评估技术及管理发生了相应转变。根据工程实践,初步论述了对管道大数据基本概念和定义的理解,探索和思考了在大数据条件下由工程适用性评估替代风险评估,依靠外因评估转变为依靠内因评估,以及风险专家评估转变为数据评估等方面的内容,同时提出了需要注意的问题和发展方向。（参13）

A thought over the pipeline risk assessment methods based on Big Data

Pipeline risk assessment techniques have witnessed rapid advancement in the past few decades, but still carry on the traditional way of risk evaluation which involves risk prediction for the concerned object according to understanding on external risk based on expert＇s risk awareness, accident probability statistics, the core problem of the method is the risk assessment model, which is based on the concept of evaluating the whole system＇s risk by limited data. With the pipeline industry entering into the age of Big Data era, Big Data of pipeline has been formed under the condition of pipeline properties alignment and included in database. With the ＂sample equal to overall＂ basis in many aspects established, the risk assessment techniques and management have changed accordingly. According to engineering practice, this paper discusses the basic concept and definition of Big Data, explores and presents the possibility of replacing risk assessment with engineering applicability assessment, replacing external cause assessment with internal cause assessment, and replacing expert risk assessment with data risk assessment, and meanwhile, issues that need to be noted and developing direction are pointed out.